EDITORIAL II
Is Depression Rooted in Personality Structure?
Many of us who entered psychiatric residency 30 years ago were trained to take
it for granted that overt symptoms can be rooted in underlying personality structures. In the
DSM-II system, we made dual diagnoses using the terminology of the time, for example,
“depressive neurosis in a hysterical personality.” This approach also reflected the pervasiveness of the psychodynamic approach, which assumed that neurotic symptoms did not appear de
novo, but emerged out of the ground of personality. In other words, symptoms are not
isolated, but reflections of the uniqueness of the person experiencing them.
The introduction of DSM-III changed our diagnostic practices irrevocably. At the same time,
psychodynamic constructs declined, while biological theories came to dominate psychiatry. In
principle, the introduction of formal Axis II diagnoses was intended to encourage clinicians to
consider the personality factors in mental disorders. In practice, as treatment methods became
more and more targeted to symptoms, clinicians often ignored the influence of personality on
symptomatology. My own experience in finding subjects for research studies on personality
disorders was instructive. Most of the patients I was interested in had been given only an Axis I
diagnosis of major depression, while Axis II diagnoses were either absent or “deferred.”
Under DSM-IV, it is not unusual for clinicians to make diagnoses by counting the required 5 out
of 9 symptoms needed for a diagnosis of “major depressive episode.” This practice can
sometimes be associated with the routine prescription of antidepressants and with a failure to
consider the possibility of offering psychotherapy. What this approach to diagnosis really fails to
address is the enormous heterogeneity of patients meeting the criteria for major depression (1). If
we took personality into account, we would be in a better position to individualize treatment
choices for patients (2).
Today, we can address the relationship of personality and depression through solid empirical
research. Thus we have learned to measure personality traits in more precise ways than by
making Axis II diagnoses. The “dimensional” approach uses standardized self-report
instruments, which allow us to study personality, not as a category, but as a set of continuous
variables straddling the boundaries between normality and pathology. To understand depression,
which is associated with diagnosable personality disorders in only a minority of cases (3), we
need measures of normally varying traits that can be predisposing factors for the development of
depressive mood.
The papers being published in this issue are good examples of this approach. Enns and Cox (4)
offer a broad, scholarly review of the literature, identifying which personality dimensions are
most likely to be associated with depressive illness. The most important of the “higher order”
factors, neuroticism, can be found in most dimensional schema of personality. The construct describes a temperamentally based trait that makes people sensitive, easily upset, and “thin-skinned.” The “lower order” factors associated with depression may, however, be more clinically
significant, since clinicians may want to target excessive dependency or self-criticism for
psychotherapeutic intervention.
The paper by Zaretsky and others (5) provides an empirical test of the strength of the relationship
between these personality dimensions and clinical depression. The findings here provide only a
partial confirmation of the theory and suggest that dependency may act through an intermediate
variable: negative life events. These findings also underline another principle that is not always
well understood by clinicians, however: negative life events do not happen “out of the blue” but
often reflect the influence of problematic personality traits (6). In other words, bad things are
much more likely to happen to difficult people! Since personality traits are themselves highly
heritable (7), this is a good example of a gene; environment interaction.
In summary, a stress–diathesis model, combining the psychodynamic and biological perspectives, is consistent with much of the research on mood disorders (1). On the one hand, depression is not just a “chemical imbalance” but is rooted in the person. On the other hand, personality traits that influence the quality of an individual's environment are under partial genetic influence. Understanding the relationship between personality and depression is a good exercise in conceptualizing the complex origins of mental disorders.
References
Joel Paris, MD
Associate Editor